[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[536]
IJESRT INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH
TECHNOLOGY COMPARATIVE STUDY OF DIFFERENT CFD MODELS TO EVALUATE HEAT
TRANSFER AND FLOW PARAMETERS IN STHE
Anshul Jain*, Prof. K. K. Jain, Prof. Sudarshan Patel
(*ME student, Prof. & Head, Prof. SRIT, Jabalpur) * Mechanical Engineering Department, SRIT Jabalpur, India
ABSTRACT Computational fluid dynamics ( CFD) simulations can be very useful to investigate heat transfer and visualize the
temperature fields & fluid flow characteristics of shell and tube heat exchanger. A shell and tube heat exchanger is
modeled to find the heat transfer parameters. The heat exchanger contains tubes inside with baffle arrangement. The
flow and temperature fields are resolved using CFD package (ANSYS FLUENT). The experimental investigation
has been also performed for comparison purpose. The CFD turbulence models considered for investigation are k-
epsilon, SST, Eddy Viscosity and Laminar model. Laminar flow is consider for understanding the significance of
turbulence in the flow field. The boundary conditions taken for the computational domain are derived out of the
experimental investigation results. Transient analysis has been performed for the physical time scale of 1800
seconds. Unstructured meshing method is used to create mesh on the domain. It has been find out that k-epsilon
model came out to be the best model to predict the flow parameters, heat transfer coefficient and behavior of present
case of STHE.Reasonable agreement is found between the simulation and experimental data.
KEYWORDS: Shell & tube; Turbulence ; CFD; heat transfer parameters.
INTRODUCTION Heat exchangers are one of the most important heat transfer apparatus that are used in industries like
chemical engineering, oil refining, electric power generation etc. Shell-and-tube types of heat exchangers
(STHXs) have been commonly and most effectively used in Industries over the years. The shell-and-tube heat
exchangers are still the most common type in use. They have larger heat transfer surface area-to-volume ratios than
the most of common types of heat exchangers, and they are manufactured easily for a large variety of sizes
and flow configurations. They can operate at high pressures, and their construction facilitates disassembly for
periodic maintenance and cleaning. The shell-and-tube heat exchangers consist of a bundle of tubes enclosed
within a cylindrical shell. One fluid flows through the tubes and a second fluid flows within the space between
the tubes and the shell. Typical Shell-and-Tube heat exchanger is shown in Figure 1.1.
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[537]
Its unavoidable need has necessitated work on efficient and reliable designs leading towards optimum share in the
overall system performance. The Log Mean Temperature Difference (LMTD) method and the number of heat
transfer units (NTU) method have been used for heat exchanger design . These methods have some shortcomings
associated with them i.e. iterative in nature and need of a prototype to implement the design. Due to these reasons,
these methods are time consuming as well as expensive especially for large scale models. However, economical
access to powerful micro processors has paved the way for evolvement of Computational Fluid Dynamics (CFD)
during the design phase.( V. Kumar, S. Saini et al 2006)
CFD is a science that can be helpful for studying fluid flow, heat transfer, chemical reactions etc by solving
mathematical equations with the help of numerical analysis. It is equally helpful in designing a heat exchanger
system from troubleshooting and optimization by suggesting design modifications.
CFD employs a very simple principle of resolving the entire system in small cells or grids and applying governing
equations on these discrete elements to find numerical solutions regarding pressure distribution, temperature
gradients, flow parameters and the like in a shorter time at a lower cost because of reduced required experimental
work (Y. Wang, Q. Dong, M. Liu et al 2007)
In the present work a shell and tube heat exchanger is modeled to investigate the heat transfer parameters. The heat
exchanger contained 14tubes inside a 1025 mm long and 156 mm diameter shell with baffle arrangement. The flow
and temperature field inside the shell and tube are resolved using CFD package (ANSYS CFX 13.0). A set of CFD
simulations is performed for a single shell and tube bundle and is compared with experimental results.
The four CFD models are considered and are compared to find the best suitable model for the present case. The
modeling has been done in design modular module in the Ansys package. The meshing has been done using un-
structural Tetrahedral mesh element. And the mesh of the element fixed after grid dependency test has been found to
be 779491 elements. (ref. Appendix.A)
The results are found to be good for the turbulence model. Further the k-epsilon model predict the flow parameters
and heat transfer-coefficient more accurately then other models.
COMPUTATIONAL MODELLING
The first step of computational modeling is Geometry Modeling. It requires the geometric parameters of
the model. It is a mathematical model that requires extensive computational resources to study the
behavior of complex system by computer simulation. Instead of deriving the analytical solution of the
problem by solving complex mathematical equations, experimentation with the model is done by setting
the parameters of the system in the computer.CFD resolves the entire system in small cells and apply the
governing equations to find numerical solutions with regard to fluid flow and temperature distribution. It
creates a virtual prototypes of the system and gives the numerical solution in a shorter time and lower cost
due to reduced required experimental work. The basic approach of usind CFD are
(A)Pre-processor (B) solver (C) Post-processor.
GOVERNING EQUATIONS
The flow is governed by the continuity equation, the energy equation and Navier-Stokes momentum
equations. Transport of mass, energy and momentum occur through convective flow and diffusion of
molecules and turbulent eddies. All equations are set up over a control volume where i, j, k = 1, 2, 3
correspond to the three dimensions.
Continuity Equation
The continuity equation describes the conservation of mass and is written as in equation
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[538]
(b)Momentum Equations (Navier- Strokes Equations )
( c ) Energy Equation
Turbulence Model :
Since the flow in this study is turbulent, so turbulence effect should be considered using turbulence
modeling. The choice of turbulence is very critical in CFD simulations. However, there is no universal
criterion for selecting a turbulence model. On the basis of literature available, and for comparing the
performance of the turbulence model; three turbulence model are consider in this study. One laminar
model is also consider to understand the effects of turbulence. In this study k − ε turbulence model, k − ω
SST model, Eddy-viscosity model and laminar viscosity model are considered.
( i ) k − ε Model : The first transported variable is turbulent kinetic energy, k. The second transported
variable in this case is the turbulent dissipation, ε. There respective modeled transport equations are as
under
For k,
And for ε
The closure coefficients for k – ε models are Cµ =0.09 , Cε1 = 1.44, Cε2 = 1.92 , σk =1.00 , σε = 1.34
The physical interpretation of the ε equation is,
1. Accumulation of ε 2. Convection of ε by the mean velocity 3. Production of ε
4. Dissipation of ε 5. Diffusion of ε
(ii) k − ω SST Model- It has been a problem to accurately predict the flow separation. The modeled
equation for k is as under
and the modeled equation for ω is
Closure Coefficients for k − ω Model are α = 5/9, β= 3/40, β∗ σk= ½ , σω = ½
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[539]
(iii) Eddy-Viscosity Model : The concept behind the eddy-viscosity model are that the unknown
Reynolds stresses, a consequence from the averaging procedure, are modeled using flow parameters
(strain rate tensor & rotation tensor) and an eddy viscosity
( iv ) Laminar viscosity model : The laminar viscosity model is used for specifying the viscous
conditions of flow. It defines the laminar flow, it is based on Navier strokes equation
METHODOLOGY AND EXPERIMENTAL WORK
Methodology:
EXPERIMENT ON THE SETUP TO COLLECT READING FOR B.COND.
SELECTION OF PARAMETERS(temp, velocity etc )
MODELING OF EXPERIMENTAL SETUP IN CAD
DESCRITIZATION OF CAD MODEL
DEFINING BOUNDARY CONDITION
SELECTION OF PHYSICAL MODEL FOR PROBLEM
ANALYSIS OF STHE BY SELECTED FOUR MODELS
POST PROCESSING THE RESULTS
COMPARING THE RESULTS AND CONCLUSION
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[540]
Experimental setup description :
The Experimental observations are as follows
1. Steam pressure (gauge) (kg/cm2) = 0.25 (kg/cm2) =0.245 bar
2. Water inlet temperature = 29.5oC.
3. Water outlet temperature = 72.4oC.
4. Water flow rate = 11.5 cc/sec.
5. Steam inlet temperature = 102.1oC.
6. Steam Outlet temperature = 81.2oC.
7. Condensate volume collected = 340 ml.
8. Time of collection = 60 sec.
RESULTS AND DISCUSSION
Figure 1,2,3,4 represents the results obtained by eddy viscosity model, k-epsilon model,laminar model
and shear stress transport(SST) model respectively.
1. Logarithmic temperature distribution of heat exchanger pipe along the length.
HEAT EXCHANGER DIMENSIONS
No. Description Unit Value
1 Shell outside diameter mm 156
2 Shell inside diameter mm 150
3 Tube outer diameter mm 12
4 Tube inner diameter mm 10
5 Number of tubes 14
6 Shell/Tube length mm 1025
7 Number of baffles mm 4
8
.
Baffle spacing
mm 205
9
.
Baffle inclination 0o
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[541]
Comments :
Eddy Viscosity model and Shear stress transportation model generated the unsatisfactory results due to
violation of exit temperature criterion. Laminar viscosity model generated the realistic results but due to
turbulences encountered in steam and water particle practically this results are also not giving the
satisfactory correlation with experimental results. So the best suited criterion for this heat exchanger
problem solving is K-Epsilon model. So the best suited criterion for heat exchanger problem solving is K-
Epsilon model. Results are matched and validated by the experimental results, because this model having
a fear balance between the turbulence and shear at wall.
2 . Heat transfer coefficient of heat exchanger pipe along the length.
Comments :
Eddy Viscosity model and Shear stress transportation model generated the unsatisfactory results due to
dominancy of the turbulences criterion consideration by these two models. So the heat transfer coefficient
is vulnerable throughout the flow. In Laminar viscosity model the flow is stable and free from eddy but
not a practical case. Results are also not giving the satisfactory correlation with experimental results.So
for the mix flow (steam and water) problems of heat transfer the K-Epsilon model gives the best results
that is because, this model having a fear balance between the turbulence and shear at wall for moderate
fluid flow conditions.
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[542]
3. Steam Temperature distribution of heat exchanger pipe.
Comments: Steam Temperature distributions inside the casing of heat exchanger are shown in the
figures. The high heat transfer zones or where the turbulence involves the temperature of the steam
reduces abruptly. This abrupt behavior is undesirable, so as the K-epsilon model shows the minimum
inconsistency across the flow line.
4.Sectional view of water temperature distribution of heat exchanger pipe.
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[543]
5.Heat transfer coefficient distribution of heat exchanger pipe.
Comment :There is a pictorial representation of Heat Transfer coefficient among the all models. K-
epsilon model represent the consistent pattern of heat transfer coefficient whereas laminar. Eddy and SST
models are non-uniform and random heat transfer coefficient generators. So for the low and moderate
velocity force convective systems like our case, the K-ε Viscosity model is best suited.
6. Steam line for stream distribution of heat exchanger pipe.
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[544]
Comments: Vortecity is generated due to shear or velocity gradient among the particle and also due to
wall and fluid interface but the major role of vortex comes into the picture when the turbulent fluid
interact with baffles inside the tubes. Eddy viscosity, laminar and SST model are not able to generate the
ideal condition for problem solving, due to over-dominancy of turbulence in Eddy and SST model.
Results are best fit for K-epsilon model. So for the general purpose discreet fluid interaction problems are
solving by this particular method.
COMPARISION OF RESULTS AND DISCUSSION :
(a)Comparison of Results of Different CFD models with Experimental Results
Sr.
No Parameter
Experimental
Result
Eddy
Viscosity
Model
Laminar
viscosity
Model
k-ε
Model
Shear Stress
Transport
Model
1 Steam Inlet Temperature (°C) 102.1 102.1 102.1 102.1 102.1
2 Steam Outlet Temperature (°C) 81.2 82.8 79.1 83.2 89.9
3 Water Inlet Temperature (°C) 29.5 29.5 29.5 29.5 29.5
4 Water outlet Temperature (°C) 72.4 99.1 57.1 69.9 86.3
5 Water Wall side CHT (W/m2K) 117.072 1407 133.5 143.5 391.5
6 Steam Wall side CHT (W/m2K) 235.180 951.1 79.93 297.24 322.8
Table 5.1: Comparative study of results.
(b)Comparison of Water Outlet temperature.
The Water outlet temperature is more accurately predicted by k-ε model than the other CFD model
considered. The outlet temperature of water predicted by k-ε model has a difference of 2.5 oC or 3.5 %
with the experimental results, while the other models show greater variation. So a good agreement is
shown between experimental and simulation result by k-ε model.
Figure: Water outlet temperature graph.
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[545]
( c )Comparision of Steam outlet temperature.
The outlet temperature of steam obtained by the considered CFD models shows a close agreement with
the experimental results except for the case of shear stress transport model which over predict the result
by 10% while the other models predict the result by ±2.5%. Moreover the Eddy viscosity model and k-ε
model shows a close agreement with the experimental results. The k-ε model predict the steam outlet
temperature with the difference of 2oC or less than 2.5%.So a good agreement is shown between
experimental and simulation result by k-ε model.
Figure: Steam outlet temperature graph.
(d) Comparison of Water wall side heat transfer coefficient.
The Water wall side heat transfer coefficient obtained by the experiment and the considered CFD models
are shown in following graph. The water wall side CHT predict by Eddy Viscosity model is much higher
than the experimental value, which is not feasible at all. Also the SST model over predict the CHT with a
large difference.The laminar viscosity and k-ε model shows a fair good agreement with the experimental
value of water wall side heat transfer coefficient.But the Pictorial representation of heat transfer
coefficient by the laminar viscosity model is non uniform and it generates random heat transfer
coefficient, while the k-ε model represent consistent pattern of heat transfer coefficient, so for the low
and moderate velocity force convective system in the present case, k-ε viscosity model is best suited.
Figure: Water wall side CHT graph
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[546]
( e ) Steam Wall side heat transfer coefficient- comment.
The steam wall side heat transfer coefficient obtained by the experiment and the considered CFD models
are shown in following graph. The eddy viscosity model over predict the steam side CHT by a much
higher value than experimental result, so it is not feasible at all. More over the laminar viscosity model
under predict the steam side CHT also the laminar viscosity model is not applicable at all in such case.
The k-ε model shows a fair good agreement with the experimental value of steam side CHT. So for such
type of models k-ε model is best suited.
Figure : Steam wall side CHT graph
CONCLUSION During the CFD analysis it has been found out that laminar model totally failed to predict the flow parameters as
well as CHT parameters. And the other entire turbulence model shows significant improvement over laminar model.
Out of three turbulence model, the SST model is over predicting the turbulence & separation due to which the
thermal aspects of the flow shows the higher value. And the same happens with the eddy viscosity model, which
predicts the high eddy formation in the baffle section which leads to the higher steam side CHT as well as water side
which is not feasible at all. Further the K-epsilon is predicting the behavior moderately. It predicts the Water side
heat transfer coefficient as well as Steam side heat transfer coefficient more closely to the experimental values. It
also predict the flow parameters, water outlet temperature, steam outlet temperature more closely to the
experimental results. It shows good agreement with the literature available on the use of k-epsilon model.
NOMENCLATURE
A = Heat transfer area (m2
)
hm = Kine t ic E ner gy
hT = Thermal Energy
hC = Chemical Energy
h = Total energy
hh = Hot side heat transfer coefficient (W/m2.K )
hc = Cold side heat transfer coefficient (W/m2
.K )
K = Exchanger wall material thermal conductivity (W/m.K )
Q = Heat transfer rate (W)
Rf = Fouling coefficient (W/m2
.K )
T = Temperature (K)
t = Time (s)
U = Overall heat transfer coefficient (W/m2
.K )
u =Velocity (m/s)
v =Fluid flow velocity (m/s)
[Jain, 4(6): June, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology
[547]
V =Volume (m3)
∆TLM
=Logarithmic mean temperature difference (K)
∆x = Exchanger tube wall thickness (m)
Φ = Potential Energy
k= turbulence kinetic energy
ε = turbulence dissipation rate.
ω = specific dissipation rate
τ = time constant for turbulence.
Cµ , Cε1, Cε2, σk , σε = Closure Coefficients for k − ε Model.
α, β, β∗ σk, σω = Closure Coefficients for k − ω Model
µt = effective turbulent viscosity
ρ =Density (kg/m3)
x, y, z = Spatial coordinates (m)
REFERENCES 1. Ahmed F. Khudheyer and Mahmoud Sh. Mahmoud 2011. Numerical analysis of fin tube heat exchanger
by using cfd technique. ARPN journal of Engineering and applied sciences, vol 6 p-7 2011.
2. Savitri Patel, D.s Patel “Computational modeling of STHE” International Journal of Engineering science
& Innovative Technology, vol.3 Issue 2 March 2014.
3. Ozden, e. & Tari, I. 2010. Shell side CFD analysis of a small shell-and-tube heat exchanger. Energy
conversion and management, 51(5):1004-1014.
4. Yongqing, W., Xin, G., KE, W. & Qiwu, D. 2011. Numerical investigation of shell-side characteristics
of H-shape baffle heat exchanger. Procedia engineering, 18:53-58.
5. Dong, Q.W., Wang, Y.Q. & Liu, M.S. 2008. Numerical and experimental investigation of shellside
characteristics for RODbaffle heat exchangers. Applied thermal engineering,28:651-660.
6. Brennan, M.S. 2003. CFD simulations of gravity sluices. (In Proceedings of the Third International
Conference on CFD in the Minerals and Process Industries held in Melbourne, Australia from 10 to 12
December 2003. Commonwealth Scientific and Industrial Research Organisation. Melbourne.p. 275-280.)
7. Tsai, Y.C., Liu, F.B. & Shen, P.T. 2009.Investigations of the pressure drop and flow distribution in a
chevron-type plate heat exchanger.International communications in heat and mass transfer,36(6):574-
578.
8. Apu Roy, D.H.Das, CFD Analysis Of A Shell And Finned Tube Heat Exchanger For Waste Heat
Recovery Applications, National Institute Of Technology, 2011.
9. Yonghua You, Aiwu Fan , Suyi Huang, Wei Liu 2011. Numerical modeling and experimental
validation of heat transfer and flow resistance on the shell side of a shell-and-tube heat exchanger
with flower baffles
10. A. M. Prithiviraj M, “Shell and tube heat exchangers. part 1: foundation and fluid mechan- ics,” Heat
Transfer, p. 33:799–816., 1998.
11. D. Kern, Process Heat Transfer. McGraw-Hill, 1950.
12. N. B. M. Bhutta, M.A.A Hayat, “Cfd applications in various heat exchangers design: A review,”
Applied Thermal Engineering, vol. 32, pp. 1–12, 2011.
13. Hetal kotwal, D.S patel, “ Cfd analysis of shell and tube heat exchanger – A review”IJESIT, vol.,
marc2013.
14. E. Cao, Heat Transfer in Process Engineering. McGraw Hill, 2009
15. B. Andersson, R. Andersson, L. Hakansson, M. Mortensen, R. Sudiyo, and B. V. Wachem, Computational
Fluid Dynamics for Chemical Engineers. sixth ed., 2010.
16. H. K. Versteeg and M. W., An Introduction to Computational Fluid Dynamics, The Finite Volume
Method. Pearson Education Limited, 2007.
17. Domkundwar & Domkundwar ., Heat and Mass Transfer data book, third ed revised,2012.
18. R. Yadav., Heat and Mass Transfer. Central Publishing House. Sixth ed., 2010
19. Muhammad Mahmood Aslam Bhutta, Nasir Hayat1, Muhammad Hassan Bashir, Ahmer Rais Khan,
Kanwar Naveed Ahmad, Sarfaraz Khan. CFD applications in various heat exchangers design: A review.
Applied Thermal Engineering 32 (2012).
20. “Ansys fluent theory guide.” http://www.ansys.com,.